Proton transport (PT) across cell membranes is a fundamental process and a
key step in many biological functions, including cell signalling and enzymatic reactions.
All biochemical reactions that convert energy from one form to another are mediated
by PT, which also serves as a vital route to achieve cell pH stabilisation. The coding for
membrane -bound proteins constitutes 25 -30% of all genes, and they are implicated in
many diseases such as diabetes and Parkinson's. Consequently, they are the subject of
major drug target studies (in fact the drug targets for all neurological diseases are
membrane -bound proteins). Whilst PT is known to occur via transient water molecules
across the cell membrane itself, it is more often the case that the mechanism involves
proteins that span the membrane surface and act as proton- specific ion channels. PT
has been widely studied in protein systems such as gramicidin A, cytochrome C oxidase,
the M2 channel protein in the influenza A virus and bacteriorhodopsin. Evidence for the
relay of H+ by buried water molecules ('water wires') mediated by the side -chains of
alpha -helices have been substantiated in these and other proteins, but finding direct
experimental evidence for the reaction pathway is extremely challenging work.
When experiment can provide only partial answers, it is the role of
computational modelling to complete the picture. Modelling these trans -membrane
proteins at the full atomistic quantum mechanical level, however, lies beyond the
capabilities of current computational techniques, necessitating the use of simplified
models. To this end, work undertaken in this thesis has derived and tested a simplified
model that is large enough to maintain the essential tertiary structures of
transmembrane proteins, but small enough to permit full ab initio MD simulations over
long time periods to be performed. The model is based on a single helix scaffold placed
under periodic boundary conditions to create a cavity that supports a water wire. The
simulations then focus on monitoring the behaviour of a proton as it 'hops' along this
wire in a manner akin to the classical Grotthuss mechanism.
Mechanistic studies have taken place using poly-glycine, poly-glycine-serine
and poly-glycine-aspartic models, and show that the mechanism of PT in channel
environments shares some features with the simulations reported for bulk water, with,
e.g., the hydrogen bond distance shortening in the time period leading up to successful
proton transfer. There are, however, also some important differences, such as the
observation of a heightened number of proton rattling events. The channel
environment also removes the need for the loss of a water molecule from the inner
coordination sphere of the receiving water molecule as the constriction in space only
allows a coordination sphere of three molecules, as opposed to four for bulk water.
The effect of varying the density of water molecules in the channel has also
been investigated. A range of cationic states have been identified, with widely varying
lifetimes and compared across all models. We also observe that the helix plays an
important role in directing the behaviour of the water wire: the most active proton
transport regions of the water -wire are found in areas where the helix is most tightly
coiled. Finally, we report on the effects of different DFT functionals to model a water - wire using the simplest poly -glycine model, and on the importance of including
dispersion corrections to stabilize the helical structure.
Finally, using the poly -glycine- aspartic acid model, a study was undertaken that
focused on the direction of proton transport through the channel when the side chains
of the aspartic acid residues interacted directly with the water wire. In this model there
were two different pathways for the excess proton to pass along: a long hydrogen - bonded network of water molecules and amino acid residues, or a short [H30]+
diffusion pathway. It was found that the proton- hopping route over multiple water
molecules and amino acid residues was preferred over the diffusion route, even though
this pathway was substantially longer.